[0001] The present invention is concerned with a method for the measurement of capacitances,
in particular of low capacitances, in which method a measurement oscillator is used,
whose output frequency is a function of the capacitance to be connected to the input
terminals of the circuit determining the frequency of the said oscillator and in which
method a known reference capacitance is used, which is connected, being alternatingly
exchanged with the capacitance to be measured, to the input terminals of the measurement
oscillator while making use of a switching arrangement.
[0002] One starting point for the present invention has been the prior-art technology that
comes out, e.g., from the FI Patents 54,664 and 57,319 (corresponding US Patents 4,295,090
and 4,295,091). In the said patents, a method is suggested for the measurement of
low capacitances.
[0003] In radiosondes, for the measurement of various parameters, in particular of pressure,
temperature and humidity, capacitive detectors are used, the magnitude of whose capacitance
depends on the parameter being measured. The capacitances of these detectors are often
relatively low, from a few pF to some dozens of pF, at the maximum about 100 pF. The
measurement of low capacitances is problematic, e.g., owing to stray capacitances,
variations in supply voltage, and other disturbances. Moreover, the said detectors
are to some extent individual, so that they have an individual non-linearity and dependence
on temperature.
[0004] In particular in telemeter applications, when, e.g., temperature, humidity or pressure
is being measured by means of electric or mechano-electric detectors,it is common
that, in connection with the measurement electronics, one or several references are
provided which are precisely known, so that errors of the measurement circuit and/or
of the detector can be eliminated.
[0005] In connection with capacitive detectors, it is known in prior art to use a reference
capacitance, which is, alternatingly with the measuring capacitance, connected to
the input of the measurement circuit, usually a circuit determining the frequency
of a RC- oscillator. By appropriately adjusting the measurement circuit or in some
other way, the corresponding output variable of the reference capacitance of the measurement
circuit can be brought to the correct level.
[0006] It is known in prior art to use measurement circuits of one reference, in particular
bridge connections, in which the measurement is, however, precise only when the electrical
value of the reference is close to the value of the detector, e.g., when the bridge
is in equilibrium. The more distant the value of the detector becomes from the reference,
the larger will also the various errors be, e.g. errors caused by changes in the dynamics
of the electronic measurement circuit. An advantage of connections with one reference
is the simplicity of the measurement circuit.
[0007] An advantage in measurement arrangements with two or more references is accuracy
of the measurement even within wide ranges of measurement, but a drawback is the complexity
of the measurement method and of the related computation.
[0008] An objective of the present invention is a further development of the prior-art measurement
circuits for low capacitances (0 to 100 pF) so that the measurement circuits become
more precise. It is an additional objective of the invention to provide such a measurement
circuit in which it is possible to eliminate the effects of switching phenomena.
[0009] A non-essential additional objective of the invention is to provide such a measurement
circuit in which the output variable is an appropriately linearized, compensated :ind
scaled DC voltage and to which, if required, a simple temperature compensation may
be connected.
[0010] In view of achieving the above objectives and those that will come out in the following,
the method in accordance with the invention is mainly characterized in that the capacitance
to be measured and the reference capacitance are connected, alternatingly one after
the other, to the input terminals of the same said measurement oscillator for the
time of an equal number of cycles of the measurement oscillator.
[0011] The frequency of the oscillator is distributed preferably by means of an asynchronous
or synchronous distributor, and the square wave of the output of the distributor is
used for controlling analog switches, which exchange the capacitance to be measured
with the reference capacitance.
[0012] In a preferred embodiment of the invention, the output of the distributor also controls
the data out buffers, one of whose operating voltage terminals has been brought out
separately. Thereby, the amplitude of the square wave supplied by the data out buffers
can be adjusted and used for linearization, compensation and scaling.
[0013] In the invention, by means of a parallel capacitance of the oscillator, a simple
temperature compensation can be provided by choosing its temperature coefficient appropriately.
The effects of the switching phenomena can be eliminated if the distribution ratio
N of the said distributor is sufficiently high, as a rule N > 10.
[0014] The temperature dependences of the oscillator affect the lengths of both of the half
cycles in the same way, and compensate most of the temperature dependence of the electronics.
The converting of the pulse ratio to DC voltage, the linearization, compensation and
scaling can be accomplished relatively simply by means of data out buffers and RC
low pass filter.
[0015] In the following, the invention will be described in detail with reference to certain
exemplifying embodiments of the invention, illustrated in the figures of the attached
drawing, the invention being by no means strictly confined to the details of the said
embodiments.
[0016]
Figure 1 shows the method in accordance with the invention as a wiring and block diagram.
Figure 2 shows the wave form of the voltage Vin, which contains the information to be measured.
Figure 3 shows, as a wiring diagram, linearization of the output voltage, advantageously
applicable in connection with the method of the invention.
Figure 4 shows different wave forms in the wiring system shown in Fig. 3.
Figure 5 shows the output voltage of the pulse-ratio-DC-voltage transformer as a function
of the capacitance to be measured.
[0017] In accordance with Fig. 1, the method in accordance with the invention is carried
into effect by making use of an oscillator 10, whereby the capacitance C to be measured
and a precisely known reference capacitance C
R are alternatingly connected between the input terminals a and b of the circuit that
determines the frequency of the said oscillator 10. Between the inlet terminals of
the oscillator 10, a parallel capacitance C
0 is connected, with which the reference capacitance C
R and C are alternatingly connected in parallel. From the oscillator 10, a frequency
f is obtained as an output variable, which said frequency f is a function of the capacitance
C
in connected to the inlet terminals:


wherein A and B are known constants, and
T = a known time constant;

[0018] The frequency obtained from the output terminal c of the oscillator 10 is passed
to the distributor 11, which is an asynchronous or synchronous distributor, whose
distribution ratio is N. From the distributor 11 the output voltage V
in is obtained, whose wave form comes out from Fig. 2.
[0019] It is an essential feature of the invention that the capacitance C
M to be measured and the reference capacitance C
R are alternatingly connected to the same oscillator 10 expressly for an equal number
of cycles. This has been accomplished by dividing the frequency f of the oscillator
10 by the distributor 11 mentioned above, whose distribution ratio is denoted with
N, and by, by means of the square wave V
in of the output of the distributor 11, controlling the switches k
1 and k
2 via an inverter 12. The switches k
1 and k
2 are accomplished, e.g., as analog switches, which are illustrated by the blocks 13a
and 13b. The switches k
1 and k
2 operate alternatingly so that when the switch k
1 is closed, the switch k
2 is open, and vice versa.
[0020] In accordance with Fi
g. 1, the output voltage V. of the distributor 11 also controls the data out in buffers
15a and 15b, one e (V
1) of whose operating voltage terminals has been brought out separately. Thereby, the
amplitude of the square wave supplied by the data out buffers 15a and 15b can be adjusted
and used for linearization, compensation and scaling, as will come out in more detail
in the following in connection with the description of Figures 3, 4 and 5.
[0021] The information on the capacitance to be measured is contained in the half-cycle
times T
1 and T
2 of the output voltage
Vin of the distributor 11, in the way coming out from the following equations (3)to
(6).




[0022] As was already stated above, the capacitance C
M to be measured and the reference capacitance C
R are alternatingly connected to the same oscillator circuit 10 for an equal number
of cycles, the number of the said cycles being advantageously equal to the distribution
ratio N of the distributor 11. Thus, the switching times T
1 and T
2 are proportional to the number N of cycles and to the connected capacitance C
in. Under these circumstances, the half-cycle lengths T
1 and T
2 of the square wave V
in of the output of the distributor 11 are determined by the distribution ratio N and
by the capacitances C
M and C
R. The pulse ratios X
1 and X
2 obtained as output, which were defined above in the equations (5) and (6), are proportional
to the magnitude of the capacitance C
M to be measured, whereas the frequency f is inversely proportional to the said capacitance
C
M.
[0023] According to a preferred emboaiment of the invention, an output voltage illustrating
the capacitance C
M to be measured is obtained from the output terminals g
1 and/or g
2 of the data out buffer circuit or of the transformer 16, which said output voltage
is linearized, temperature-compensated and scaled, making use of the following exemplifying
solutions.
[0024] The said transformer 16 may be made of one part, in which case, out of the components
15a and 15b, it includes only the transistors M
1/M
2. Alternatively, the transformer 16 may be made of two parts, in which case, besides
the said transistors M
1/M
2, it also includes an inverter 14 and transistors M
3 and M
4'
[0025] Output voltage with a one-part transformer:

wherein


[0026] Correspondingly, with a two-part transformer:


[0027] By varying the voltage V
1 to be supplied through the terminal e to the transformer 16, it is possible to adjust
the amplitude of the square wave in accordance with the equation (6a) and, thereby,
to provide an advantageous compensation, linearization and scaling of the output voltage.
[0028] The starting point of the compensation is that, in accordance with the above formulae
(4a) and (6a), the output voltages either V
01 or V
01-V
02 can be compensated in respect of the desired parameter, e.g. the temperature, by
arranging the voltage V
1 as a feedback voltage and the latter as in a suitable way dependent on the parameter
to be compensated.
[0029] The main principle of linearization is that the output voltage is feedback-connected
via the terminal e to the transformer 16 as a feedback voltage V
1. Thereby, the switch 15a/15b acts as a non-linear component, and as a result is obtained
non-linear dependence of the output voltage on the pulse ratios X
1 and X
2 defined above (equations (5) and (6)).
[0030] The main principle of scaling is that, by adjusting the voltage V
1, in the method in accordance with the invention, the range of variation of the output
voltage, constituting the output variable dependent on the capacitance C
M to be measured (equations 4a and/or 6a), i.e. the scale, can be set at a suitable
level.
[0031] In the following, with reference to Figures 3, 4 and 5, a more detailed exemplifying
embodiment of the linearization, temperature compensation and scaling of the output
voltage used in connection with the method in accordance with the invention will be
described.
[0032] In accordance with Fig. 3, the input of an operation amplifier 17, whose amplification
is denoted with G, is connected to the output terminals g
1 and
g2 of the transformer 16 described above. The output of the operation amplifier 17 is
connected to the output of a resistively feedback-connected operation amplifier 18.
From the operation amplifier 18, the output voltage V
2 is obtained.
[0033] The various wave forms V
in, V
in' and V
in" of the connections illustrated in Fig. 3 are shown in Fig.4. With different parameters
of the wiring system of Fig. 3, the following equations apply, the X
1 and X
2 present in them being defined in the above equations (5) and (6).






whereby, as fitted into equation (9),


and, fitted into equation (8), we obtain

[0034] When the equation (15) is examined, it can be noticed that by changing the sign of
the amplification G of the operation amplifier 17, the direction of the curvature
of the characteristic curve of the circuit can be changed, and by means of the term
k
2 (equation 12), the magnitude of the curvature can be adjusted.
[0035] In the following, with reference to Fig. 5, the use of a parallel capacitance C
0 for temperature compensation of the circuit will be described.
[0036] Frequency of a RC oscillator:

[0037] Temperature has a typical effect on the terms B and . B includes the temperature
dependences of DC voltage and low frequency as well as the temperature dependences
of the delay of .
[0038] On the basis of (1), (3) and (4), the lengths T
1 and T
2 of the half cycles of the pulse ratio output V. are: in


[0039] In accordance with the above equation (7), the output of the pulse-ratio-DC-voltage
transformer 16 is:

[0040] By appropriate choice of the temperature coefficient of C
0 (1), it is possible to compensate the temperature dependence of 2
T/B partly.
[0041] With reference to Fig. 5, it is noticed that the temperature dependence of 2t/B attempts
to alter the angle factor in Fig. 5 at high operating frequencies, whereby the share
of τ in the time of the half-cycle is important. This can be compensated by means
of C .
[0042] In the following, the patent claims will be given, whereby the various details of
the invention may show variation and differ from the details described above, within
the scope of the inventive idea defined in the said claims.
1. Method for the measurement of capacitances, in particular of low capacitances,
in which method a measurement oscillator (10) is used, whose output frequency (f)
is a function (f = F(Cin)) of the capacitance (Cin) to be connected to the input terminals (a,b) of the circuit determining the frequency
of the said oscillator and in which method a known reference capacitance (CR) is used, which is connected, being alternatingly exchanged with the capacitance
(CM) to be measured, to the input terminals (a,b) of the measurement oscillator (10)
while making use of a switching arrangement, characterized in that the capacitance
(CM) to be measured and the reference capacitance (CR) are connected, alternatingly one after the other, to the input terminals (a,b) of
the same said measurement oscillator (10) for the time (T1 and T2) of an equal number (N) of cycles (T = 1/f) of the measurement oscillator (10).
2. Method as claimed in claim 1, characterized in that the frequency (f) dependent
on the capacitances (CO,CR,CM) connected to the input terminals (a,b) of the measurement oscillator (10) is passed
to a distributor (11), whose distribution ratio (N) directly determines the number
(N) of the said switching cycles.
3. Method as claimed in claim 2, characterized in that the square wave (Vin) obtained from the output (d) of the said distributor (11) is used for controlling
the switches (kl,k2) which, while alternatingly exchanging them, connect the known reference capacitance
(CR) and the capacitance (CM) to be measured, each in its turn, to the input terminals (a,b) of the measurement
oscillator (10), so that an exchange of the level (Vdd,0) of the said square wave (Vin) produces an operation of the said switches (k1,k2) in such a way that the capacitance (CM) that is the next one in the sequence and that is to be measured is connected in
place of the reference capacitance (CR), and vice versa.
4. Method as claimed in any of the claims 1 to 3, characterized in that the connection
times (T1 and T2) of the various capacitances (CR and CM) are proportional to the number (N) of the connection cycles and to the connected
capacitance (CM or CR) .
5. Method as claimed in any of the claims 2 to 4, characterized in that, in order
to eliminate the effect of switching phenomena, the said distribution ratio (N) has
been chosen as N > 10.
6. Method as claimed in any of the claims 1 to 5, characterized in that in the method
a transformer (16) is applied, as whose input signal the square wave (Vin) obtained from the said distributor (11) is passed and from whose output terminals
(g1 and/or g2) a DC voltage representing the capacitance to be measured is obtained, which said
DC voltage is linearized, temperature-compensated and/or scaled.
7. Method as claimed in claim 6, characterized in that a feedback voltage (V1) is fed through a certain terminal (e) of the said transformer (16), the amplitude
of the said square wave (V1) being adjusted,by varying the said feedback voltage, for the purpose of compensation,
linearization and/or scaling of the output voltage of the transformer.
8. Method as claimed in claim 7, characterized in that, for the purpose of linearization
of the output voltage of the said transformer (16), the said output voltage is passed
as the said feedback voltage (V1) to the non-linear component of the transformer so that a controlled non-linear dependence
of the output voltage on the pulse ratios (X1,X2) of the said square wave (equations (5) and (6)) is produced.
9. Method as claimed in any of the claims 7 or 8, characterized in that in order to
scale the output variable, the level of the said feedback voltage (V1) is adjusted.
10. Method as claimed in any of the claims 6 to 8, characterized in that for the temperature
compensation of the measurement circuit, a parallel capacitance (C ) connected to
the input terminals (a,b) of the measurement oscillator (10) is used, the temperature
coefficient of the said parallel capacitance (C ) being fitted so that the temperature
dependences of the oscillator (10) are at least partly compensated (Fig. 5).
11. The use of a method as claimed in any of the claims 1 to 9 in radiosondes, in
the telemeter measurement of pressure, temperature and/or humidity.